[Previous] Corroboration | Home | [Next] Handling Information Overload

Critical Preferences

Originally posted Feb 2010 at http://www.criticalrationalism.net/2010/02/28/critical-preferences/

What problem is the idea of a “critical preference” intended to solve? (And how does it solve it?) I think the problem is this:
We form theories to solve our problems, and we criticize them. Sometimes we decisively refute a theory. If this leaves us with exactly one theory standing, there is no problem, we should prefer that theory.

Refutations can be hard to create. Often there are several theories offered as solutions to one problem, which contradict each other, but which are not decisively refuted. What are we to do then? The intellectual answer is to invent new criticisms. That may take years, so there is a pragmatic problem: we must get on with our life, and some of our decisions, today, may depend on these theories.

The idea of a critical preference is aimed to solve the pragmatic problem: how should we proceed while there is a pending conflict between non-refuted theories?
Popper proposes (without using the term “critical preference”) that we can form a critical preference for one theory, and proceed using that theory in preference to the others. The critical preference should be for whichever theory best stands up to criticism, or in Popper’s words the theory that “in the light of criticism, appears to be better than its competitors” (C&R p 74). Popper writes something similar in Objective Knowledge, p 82 (see also pages 8, 16, 22, 41, 67, 95, 103). Similarly, Firestone wrote, “The best problem solution is the competitive alternative that best survives criticism.”

(How we judge which theories are better, or best survive criticism, is another question, and Popper gives various guidance (e.g. C&R p 74, and the idea of corroboration), as does Deutsch (e.g. his recommendation to prefer explanations that are hard to vary), but I’m not interested in that here.)

Would others here agree that this is the problem and solution of critical preferences? (My purpose here is that I think it is mistaken, and I want to get the theory right prior to offering criticism. Perhaps I’ve misunderstood it.)

Follow up post: http://curi.us/1489-critical-preferences-and-strong-arguments

Elliot Temple on July 11, 2013

Messages (76)

I think Popper was wrong to suggest that the one that survives the severest criticism should be preferred. All that follows logically, given the premise that our preference is for the truth, is that those that have been criticised should not be preferred. it follows then that we can be agnostic about the last surviving solution, given the state of criticism, it does not follow that it should be accepted. To claim otherwise is called the eliminativist fallacy. You can accept it if you choose, but your acceptance does not follow and is not determined by the investigation. Since an investigation can only give negative advice. That is the lesson of critical rationalism.


Anonymous at 2:33 PM on July 18, 2017 | #8841 | reply | quote

if the last surviving idea says "you should do X now" and you have no criticism of it, then you should do X now.

if the last surviving idea says "you can solve Y problem by Z action", and you have no criticism of that, and you want to solve Y problem, then you should do Z action.

> That is the lesson of critical rationalism.

no, it's *your* position. it's not CR's position.


curi at 3:32 PM on July 18, 2017 | #8842 | reply | quote

or if the last surviving idea says you should believe X about Y, and you have no criticism of that (any doubts, hesitations, objections, any reason that'd be a bad idea, any guess that's mistake, any lead that there might be a superior alternative, etc, etc), then you should believe X about Y.


curi at 3:34 PM on July 18, 2017 | #8843 | reply | quote

> no, it's *your* position. it's not CR's position.

Err, what? I said it was a lesson of CR. I never spoke about positions.

it does not follow from something having no criticism that you should do it. This is just your own proposal and it does not follow from the investigation, which was exactly my point.


Anonymous at 4:35 PM on July 18, 2017 | #8844 | reply | quote

it's not a lesson of CR, it contradicts CR.

> it does not follow from something having no criticism that you should do it.

call the idea X.

it does follow (that you should do X) if X says you should do X (and you have no criticism of X). you need to pay more attention to what X is, which is what i was explaining and you didn't engage with (you just repeated the same mistake i'd just refuted with 3 examples, and responded to none of the 3).


curi at 4:38 PM on July 18, 2017 | #8845 | reply | quote

Let me be more precise in my statement. Just because the statement "you should do x now" has no criticism, does not mean you should accept the statement. That is justificationism.


Anonymous at 4:40 PM on July 18, 2017 | #8847 | reply | quote

if you have no criticism of accepting the statement, and it says you should accept it, then how would it make any sense not to accept it? you have zero objections to accepting it. you seem to just want life and thinking to be arbitrary whim.


Anonymous at 4:42 PM on July 18, 2017 | #8849 | reply | quote

not accepting a theory, T, is itself an action/choice/option. you have a criticism of not accepting T because the premise was that accepting T was the ONLY option/theory/whatever that survived criticism.

if only one option survives criticism, there are NO ALTERNATIVES that have survived criticism, including "do nothing".


Anonymous at 4:44 PM on July 18, 2017 | #8850 | reply | quote

If there are many proposals including the proposal to do nothing, and all of the proposals have been criticised apart from one. Then you should accept the one that's left, since there are no other proposals that one can accept. But this is false, because all of the proposals can be accepted whether or not they have criticisms, since a criticism is itself sometching we accept as being successful. The criticism is false relative to the proposal and the proposal is false relative to the criticism, it requires a decision. It does not follow form the investigation that I should accept any of them.

If I accept the state of the investigation, then it follows that there is only one options. But I might not accept that I was rigorous enough.


Anonymous at 5:32 PM on July 18, 2017 | #8855 | reply | quote

>It does not follow form the investigation that I should accept any of them.

This refers to the criticisms of the proposals not to the proposals.


Anonymous at 5:34 PM on July 18, 2017 | #8856 | reply | quote

your new take is maybe you decide some other ideas have survived criticism after all, so we don't have the situation of having exactly one theory which survived criticism (which is the situation we were talking about). i don't get the point. we were talking about whether, having decided there is exactly one idea which survived criticism, you should therefore accept it.

and if you don't think the investigation was rigorous enough, that is a criticism of accepting the one idea remaining from the investigation. so you, again, don't actually have the situation being discussed which is exactly one idea which survived criticism.


Anonymous at 9:05 PM on July 18, 2017 | #8860 | reply | quote

You added into the discussion that the one surviving solution to the problem. Includes the idea that doing nothi has also been criticised, but doing nothing is not a solution.

Now we have to break this down into two different scenarios.

The first scenario is that of accepting a scientific theory. The fact that you have one surviving theory, it does not follow that you should accept it. You can remain indifferent.

The second scenario is one in which you in a situation where you have to act in some way to solve a problem and the idea is that there is one solution left and that doing nothing has been criticised. But this leads to a serious problem. And that is if it turns out that you have a criticism of this solution. This puts one in an awkward position. Since all possible proposals for actions have been criticised.


Anonymous at 10:05 PM on July 18, 2017 | #8861 | reply | quote

>and if you don't think the investigation was rigorous enough, that is a criticism of accepting the one idea remaining from the investigation.

No, it is not because accepting an idea is to do with whether it is true. A criticism is directed at its truth. That the investigation was not rigorous enough is not a criticism of the proposal.


Anonymous at 10:09 PM on July 18, 2017 | #8862 | reply | quote

> but doing nothing is not a solution.

of course it is. it's an action you can take in regards to a problem. it may solve your problem, or not.

> The first scenario is that of accepting a scientific theory. The fact that you have one surviving theory, it does not follow that you should accept it. You can remain indifferent.

no, because being indifferent is a rival theory which is refuted (by premise of only one surviving theory).

> Since all possible proposals for actions have been criticised.

that's a COMMON situation. you then have to brainstorm new solutions.

if action becomes URGENT then you brainstorm solutions to the NEW PROBLEM of "what should I do about X situation, given that I am under this time pressure?" the new problems get easier as one more urgently needs a solution and therefore can come up with solutions like "X isn't ideal but b/c i have so little time left I'll take it." (in that way, and X you criticized previously may solve the new, easier problem.)

this is explained at e.g. http://fallibleideas.com/avoiding-coercion


Anonymous at 10:14 PM on July 18, 2017 | #8863 | reply | quote

> No, it is not because accepting an idea is to do with whether it is true. A criticism is directed at its truth. That the investigation was not rigorous enough is not a criticism of the proposal.

accepting an idea is an action, a human decision. whether or not you've investigated something enough to accept it, or you should be undecided for now and think about it more first, is relevant to human actions like this.


Anonymous at 10:15 PM on July 18, 2017 | #8864 | reply | quote

given some situation, some evidence, some known candidate solutions, etc, etc, there is a right answer (and objective truth) about what you should do about it and what judgement you should reach.

if you have exactly one non-refuted theory about what this objective truth is, and no criticisms of it, you should proceed accordingly. what else could you do? all alternatives are refuted.

your attempts to get around this often rely on using terminology which is reasonably understandable in general conversation, but not suited to speaking precisely about epistemology.


Anonymous at 10:17 PM on July 18, 2017 | #8865 | reply | quote

> Since all possible proposals for actions have been criticised.

that's a COMMON situation. you then have to brainstorm new solutions.

That is a proposal for action. What if you cannot come up with any proposals that work?


Anonymous at 11:46 PM on July 18, 2017 | #8866 | reply | quote

> That is a proposal for action. What if you cannot come up with any proposals that work?

see my essays on the matter, e.g. the linked one above and others at http://curi.us/1595-rationally-resolving-conflicts-of-ideas


Anonymous at 11:54 PM on July 18, 2017 | #8868 | reply | quote

>> but doing nothing is not a solution.

>of course it is. it's an action you can take in regards to a problem. it may solve your problem, or not.

This reply only makes sense as a reply if you substitue "action" for "solution" in the original claim.

The original claim points out that a proposal to not implement a solution, ie by doing another action, is not itself a solution to the problem.

The whole point was that you can be indifferent to the last remaining solution. Indifference is an action. It does not follow automatically that it is therefore a solution to the problem.

Maybe this is where the miscommunication entered.


Anonymous at 3:47 PM on July 11, 2018 | #10037 | reply | quote

>given some situation, some evidence, some known candidate solutions, etc, etc, there is a right answer (and objective truth) about what you should do about it and what judgement you should reach.

Yes, but it cannot be determined. You seem to assume that because there is such a thing that it can be determined. You should do what's right, but determining what's right can only be approached negatively, and that can only take you so far before you have make a choice. Positive information about any situation is pure guesswork.


Anonymous at 3:55 PM on July 11, 2018 | #10038 | reply | quote

#10037

doing nothing is a *candidate solution*, rather than a different kind of thing that can't be considered as a (potential) solution.

> The whole point was that you can be indifferent to the last remaining solution.

"indifferent" is ambiguous about whether you think it's a solution, or not, and whether you have any criticism of accepting it and proceeding accordingly, or not.

> Yes, but it cannot be determined.

you can create knowledge about it. dunno why you're denying this. i guess because you don't see the role of criticism and CR in reaching positive answers to anything. well, why come here and assert that ET, KP, DD are wrong, indirectly, without addressing their published explanations? did you not realize people would disagree with you? you thought the response would be agreement that we just can't figure things out like in the quote you're arguing against? (it's weird to quote something, argue with it, then act like you except agreement)


Anonymous at 4:01 PM on July 11, 2018 | #10041 | reply | quote

>you can create knowledge about it.

I never denied this. But knowledge is created through guesswork. Guesswork can be true or not true.

CR proposes guesswork and criticism. Guesswork gives us the material, criticism wittle's it away.

I come here to offer criticism.

Your hostility is breathtaking.


Anonymous at 4:10 PM on July 11, 2018 | #10043 | reply | quote

what does

> cannot be determined

mean if not that we can't know?

are you say we *can* know, but can't determine it?

you are being unclear. i'm not hostile. why did you think i am? you didn't quote anything and explain any hostility in it.


Anonymous at 4:12 PM on July 11, 2018 | #10044 | reply | quote

Yes, we can conjecture the truth.


Anonymous at 4:14 PM on July 11, 2018 | #10045 | reply | quote

you aren't clarifying.

can you introduce yourself? have you read ET, KP, DD? we don't seem to have similar background knowledge.

i guess you stopped trying because you think i'm hostile, so now we won't be able to get anywhere. you didn't really answer any of my questions, nor did you state some kind of overall goal or purpose. and your last comment didn't try to lead anywhere – it had no questions and no new directions for conversation. so i don't know where to go from here.


Anonymous at 4:16 PM on July 11, 2018 | #10046 | reply | quote

Wny not engage with what I am saying, instead of trying to find out my credentials.

If we have a set of proposals, we can determined, by criticism and evidence, that they are in error, fallibly of course, but you can't determine that they are not in error in the same way, we can only conjecture that they aren't So if we have one left over it does not follow that it does not contain error, we can conjecture it does not, but that is not determined by evidence or criticism. We can have the truth but never know it.


Anonymous at 4:23 PM on July 11, 2018 | #10047 | reply | quote

> Wny not engage with what I am saying

i'm trying to. i tried to get clarification. you still have not clarified what you think "determine" means and how it differs from "know". we seem to be so far apart in perspective it's hard to understand each other, so i asked about that (it wasn't a credentials check). when people seem to not understand my clarifying questions, at all, i start looking for points in common, and to get some sense of what background knowledge i can and can't refer to (since my defaults for those things weren't working). i was trying to fix communication problems – while also continuing to engage with the topic. but i don't really know what you're trying to say or what you want.

> but you can't determine that they are not in error in the same way,

who said anything about determining something is "not in error"? here you seem to use "determine" as a synonym for "know", contrary to before, and without directly saying what you mean.

and you seem to mean we can't even fallibly know that something is the *final truth* (no errors will ever be discovered in it). you seem to keep going back and forth between making veiled references to omniscience, or not, and it's confusing. you

seem to be trying to draw some distinction between positive and negative knowledge, but it's not clear. is it your own idea? if so, you'll have to explain more fully. or are you trying to refer to something i've already read somewhere (e.g. one of KP's ideas), and i maybe do know what you mean but i'm not matching your words to KP's words?

i do get "We can have the truth but never know it." is a KP reference but it doesn't clarify what your point is for me, nor what this has to do with the pre-existing discussion here.


Anonymous at 4:31 PM on July 11, 2018 | #10048 | reply | quote

Conjectures are either consistent or inconsistent with other conjectures. If we have an inconsistency it points to an error. This is something negative. If there is consistency we don't learn anything that is not already in the conjectures ithemselves. What gives us the positive, ie claims about the world, is the conjecture itself.


Anonymous at 4:53 PM on July 11, 2018 | #10049 | reply | quote

This doesn't help me. You aren't addressing the terminology issues or clarifying your previous statements. And you're introducing new issues and concepts, and referring to a great deal of background knowledge which seems to be partially Popperian but which you are unwilling to talk about. If you won't discuss your background knowledge then you need to include more explanation in your statements because it's much harder for me to fill in the blanks for you.

Is #11049 supposed to help me know how you use the word "determine"? Is it arguing with something by anyone?

I think it's false that a consistency check between conjectures A and B is already contained in their conjunction. A consistency check is a substantive thing in its own right. But this is another can of worms which is kinda tricky and I don't think we could talk about it when we're still stuck on "determine" and you seem basically unwilling to answer direct questions.

(I'm the same person you were talking to in the other anonymous messages in this topic today. Two anonymous people is confusing so I signed this time.)


Dagny at 4:58 PM on July 11, 2018 | #10050 | reply | quote

"A consistency check is a substantive thing in its own right."

The failure of finitistic program in mathematics shows that you cannot prove, in the mathematical sense, that something is consistent (it's consistency always remains conjectural) only that your check did not reveal an inconsistency. This procedure is unending and is only stopped by a decision to either remain indifferent or to keep the conjecture that it is consistent. Saying that it is substantive is claiming that your decision to stop at some point tells us something about the theory. Which it does not.


Anonymous at 5:06 PM on July 11, 2018 | #10051 | reply | quote

To the anon who wrote #10037

You're making many different claims. Is there a particular one you think is important and want to focus on? And did you want to talk with me, was that one of your goals? I'm open to discussion if that's what you're looking for, but what you guys are doing is too confusing. I think even your first two comments, before the other person talked, brought up too many different issues without enough explanation.

If you'd like to refocus on one thing to start with – especially if you give an idea and explain it including explicitly saying what problem it's trying to address and how it addresses it – then I could try to comment. It'd also be helpful if you said what you're looking for in this discussion (your goals, including discussion length like if it's just some entertainment today or there are longer term plans) so I could evaluate if I could help with them or not, and if I'd want to.

Sorry for the lack of epistemology in this comment (I noticed your aversion to meta discussion) but I think it'd be significantly better if you picked an issue rather than me picking.


curi at 5:08 PM on July 11, 2018 | #10052 | reply | quote

> If there is consistency we don't learn anything that is not already in the conjectures ithemselves.

You're trying to correct me on math which I wasn't talking about. We don't normally evaluate our ideas mathematically.

We do learn things when we look for inconsistencies, even when we don't find any so far. E.g. there are plenty of times it's totally non-obvious how two ideas connect or interact, and figuring that out is an accomplishment (even if we find out they're compatible).

You're adding more new concepts, tangents and complexity to the discussion before making any progress on resolving the earlier difficulties. And you're in lecture mode instead of a mode where you answer any questions (e.g. what you mean by "determine" or "know"?).


Dagny at 5:14 PM on July 11, 2018 | #10053 | reply | quote

I am not saying we don't learn things through checking. But we learn in only two ways, conjecture and criticism. When we are looking over a theory or set of theories we have expectations and those expectations are corrected or nothing happens. When you learn something about how two theories connect you are making a conjecture, and then something might grab you because you have an expectation that is challenged by their conjunction, and that is, I conjecture, due to the fact that something in your background knowledge was inconsistent with both of those two theories being true, but not with one individually and it reveals a problem or corrects your expectations. The other is that you had an unexplained thing, which those two things connected explained. So trying to find out where theories are consistent, will only teach us something if we have background knowledge that was either unexplained or inconsistent with the theories.

Another way of looking at this. you might have two expectations that you conjectured at different times, but they never came into conflict, but you might come across something that brings the two expectations into conflict.


Anonymous at 5:40 PM on July 11, 2018 | #10054 | reply | quote

> I am not saying we don't learn things through checking [for consistency].

But that was what you started debating me about...

I criticized your:

>>> If there is consistency we don't learn anything

and so on.

You seem dead set on providing mini lectures but not answering questions. And you present yourself at knowledgeable about CR. I don't know what to do with that. Since you don't want to learn from me, I'd ask you to teach me something that I don't know (as against the things you've been talking about which are a mix of mistakes and things I know in much more detail than you've gone into), to see if you could do that. I do have questions about CR, points of curiosity. But that would be a question. So idk. Now what? Help?


Dagny at 5:46 PM on July 11, 2018 | #10055 | reply | quote

By determine, I mean that when we investigate statements for their truth value, what does the investigation allow us to conclude logically about the relationship between the evidence and criticism, and the theory (or proposal) . Positivists claim that evidence can tell us that the theory is true, or probably true, and that the more criticisms it overcomes the more likely it is to be true. Negativists claim that evidence and criticism, at most, rule out theories or proposals and don't make them any more likely to have been true or false, on the basis of the evidence. Derermination is what is logically concluded about the statements in play in regards their truth or falsity.

When you claim that a consistency check can be substantive, I think (1) You are not understanding (this might well be my fault) and (2) you are mistaken about what it is that is substantive. Usually when we are trying to find out what two theories say about the world, which if they are consistent and have strong claims, can be substantive. But you don't learn this from a consistency check. This is a content check, if you will - what do these two say about the world assuming that they are true and consistent. This is what you are mistaking for being substantive in a consistency check. Think of this barebones logically. You can transalte any theory into a formal system and take away the content, a consistency check is about its logical structure and not about the content of its claims Although you might find something out through guesswork, while doing a consistency check, if the content is left in.

If I am mistaken here, please can you explain your claim that a consistency check can be substantive.


Anonymous at 11:56 AM on July 12, 2018 | #10056 | reply | quote

> By determine, I mean that when we investigate statements for their truth value, what does the investigation allow us to conclude logically about the relationship between the evidence and criticism, and the theory (or proposal) .

that is not a definition of "determine". you didn't address "know" either.

what does it mean to "determine" something, in your terminology?

in my terminology:

determine = know = fallibly know = conjecturally know.

those are all the same thing.

from what i can tell, your usage hasn't been consistent. but it's hard to tell. maybe you could give some word meanings so that all your messages are consistent. that's what i'm asking for.

also, why are you ignoring ET for me? it's his blog, you seem to dislike me, and i thought his suggestion to pick one thing was productive. (on a related note, i'm not going to try to answer everything you said all at once. we need to slow down. if we address some things then we can get to others after. so this message intentionally tries to focus on one issue.)


Dagny at 12:02 PM on July 12, 2018 | #10057 | reply | quote

I have already explained my terminology. The very first paragraph explains it. Determine means what can be logically deduced from what, given the statements on plat. At least some guesswork at any particular time cannot be deduced from the statements in play.


Anonymous at 12:10 PM on July 12, 2018 | #10058 | reply | quote

> Determine means what can be logically deduced from what, given the statements on plat.

Are you ESL?

I don't know what "on plat" is. Double typo for "in play"?

Have you tried looking up "determine" in a dictionary?

"what can be logically deduced from what, given the statements on plat." is dissimilar to e.g. "to decide" or "ascertain or establish exactly, typically as a result of research or calculation" (both from dictionary, second one kinda infallibilist).

Your definition isn't even the right part of speech. "Determine" is a verb, but your text "what can..." is a noun phrase (it's a thing, not an action).


Dagny at 12:14 PM on July 12, 2018 | #10059 | reply | quote

I agree with grammar correction. But you asked me to explain my terminology which I did, you did not ask me to get the dictionary definition. So your slight of hand criticism is troublesome and not in good spirit.

To determine here used means to show that one thing lfollows from another thing, using the rules of logic.


Anonymous at 12:20 PM on July 12, 2018 | #10060 | reply | quote

You aren't explaining what you think "determine" means. You aren't even giving an answer in the right category. You haven't given something that even could be an answer, let alone one that addresses key issues like what knowledge is and whether fallible knowledge counts.

My conclusions:

- you are not a fluent English speaker

- your internal thinking is confused in whatever language it's in. there are translation difficulties but they aren't the core issue

- you're making tons of errors

- you're blaming others for your errors

- you think others are mistaken, when actually you're mistaken

- you are hostile to criticism

- you are hostile to discussion methodology discussion

- you won't talk about your goals, skills, plans, what you want, what you're offering, how long you will stay (in any scenarios, i know it could vary), etc

- you are hostile to meta discussion

- you get emotional about criticism, meta discussion, mistakes, etc, and the emotions affect your discussion quality. and you externalize them and start accusing others of being emotional, hostile, etc. and you're quite eager to do that, you have a thin skin and don't put much effort into giving people the benefit of the doubt (you initially got upset after the very first reply, and it's been a recurring issue, but you don't want to do problem solving regarding it either. you wouldn't say which thing(s) upset you initially, or why, or any request for me to change anything. there was no problem solving possible, by your choice.)

- you don't care much about your mistakes

- you won't go slow and careful and try to get things right

- you usually won't even try to answer questions

- there's no one important philosophical topic you care about and want to discuss or focus on, and think you have anything good to say about (a solution to a problem)

I think this is an impasse and you are blocking any paths forward. (Whether you even know what Paths Forward are is an example of the kind of question you've refused to answer, so I'm unable to know if you'll understand what I'm saying or not. I can't calibrate what I say to what you know and don't know because you refuse to help me do that.)


Dagny at 12:35 PM on July 12, 2018 | #10061 | reply | quote

> you think others are mistaken, when actually you're mistaken

This could be true. But it could easily apply to you. Isn't this an example of bad spirit? I am open to criticism.

There are lots of demands you are making and some of them are contradictory. Ie. You want me to focus on something and then when I am focusing on something, you accuse me of not dealing with other issues. I offered a pretty clear explanation. Now you are giving up because you are frustrated and are accusing me of being emotional. I wanted to come here to discuss what I think is a difference between my view of decision making and the one on ithis site.

It revolves around this

If we we have one theory left after a round of investigation, does it follow logically from the investigation that this theory that is left over is true and that it should be accepted.

My claim is that it does not follow. It could be true and it can be accepted. That's the most we can derive from the state of the investigation.


Anonymous at 1:02 PM on July 12, 2018 | #10067 | reply | quote

> I have already explained my terminology.

This is an example of one of the many ways you kept asserting the other guy was wrong, not you. You're basically saying "omg just listen better".

> I offered a pretty clear explanation.

You literally aren't fluent enough in English to understand what the word "determine" means or how to explain what you mean by a word. You're unable to answer that issue and move the discussion forward. And instead of having some humility and trying again, you blame others and make repeated remarks about how you already answered (meaning: it's their fault for not understanding what you already said – an awkward thing for a non-fluent speaker to say to a fluent speaker about word meanings!).

You aren't able to communicate, you're impatient about it and think we should be able to move on already (instead of being happy to keep at the issue until we get somewhere), and you won't admit your problems (like not being fluent enough in English) and talk about ways to address them.

And you continue to ignore relevant direct questions, e.g. why you're ignoring ET and his question, and whether you're ESL (I think I know the answer, but still I directly asked and you ignored me). Your choice to keep ignoring most of what I say, combined with your inability to communicate and your unwillingness to work on your communicate are an impasse. You also ignored everything I said about Paths Forward.


Dagny at 1:08 PM on July 12, 2018 | #10068 | reply | quote

You still haven't even tried to give a definition of "determine" which is the right part of speech. I kept asking but you won't try to do that. That's why I was giving up. You just stopped even trying to continue. I think this was clear but you weren't fluent enough to understand it, and there's no way forward given the combination of your language skills and your attitude.


Dagny at 1:10 PM on July 12, 2018 | #10069 | reply | quote

My mistake, I missed it in the attacks, typos, and bad grammar:

> To determine here used means to show that one thing lfollows from another thing, using the rules of logic.

That is the right part of speech.

But it's not what "determine" means in English. You're using words wrong and it's a major discussion problem and you're really bad at clarifying.

Further, your definition doesn't make sense when applied to some of your prior messages.

> You seem to assume that because there is such a thing that it can be determined [as a deductive implication].

Where was that assumed? It wasn't.

> If we have a set of proposals, we can determined, by criticism and evidence, that they are in error, fallibly of course,

In standard English, this is correct. But with the definition you now claim to be using, it's false. You can't deduce the falsity of theories given proposals (candidate theories to solve a problem), criticism and evidence. Also I don't know what you think "criticism" is, but you can't mean what KP/ET/DD mean (explanations of how/why ideas are mistaken) because that isn't deductive.


Dagny at 1:17 PM on July 12, 2018 | #10070 | reply | quote

I am willing to keep at this. I will try to address at least some of the errors you point out.

When I say I have been pretty clear. This is a conjecture which is not criticised by saying "you are not explaining yourself well". Instead of this I should probably say that you have yet to criticise the follow-up attempt I have made. I will try to be less flippent with my remarks. It's not good for conversation of you just you are not explaining, since I don't know what it is from that statement what is causing trouble.

Your demands about what I should do are still contradictory.

We can excise the word determine from the discussion, since I think it's causing too much trouble.

And we can say what logically follows from what in the investigation.

If you conjecture P1 and I conjecture P2 as candidates for the truth. Then we enter into a round of investigation and there is a criticism of P1 and we decide that we have to halt the investigation (at time t) . It does not follow that because P1 is criticised and P2 is not (at time t) that P2 is true (since no investigation can ever be exhaustive). Or that it should be accepted. Since (1) if you did not accept it in the first place, it would be an example of the eliminativist fallacy and (2) It commits the naturalistic fallacy (3) Claims such as "What is left over after an investigation should be accepted" is itself a conjecture that does not follow from the investigation. And is a non-sequitur. Given 1 and 2.

All of this is is compatible with the conjecture that P2 is true and also accepting P2, tentatively, of course. You don't add anything to the discussion by saying it should be accepted, other than dogmatism.


Anonymous at 1:35 PM on July 12, 2018 | #10076 | reply | quote

>Also I don't know what you think "criticism" is, but you can't mean what KP/ET/DD mean (explanations of how/why ideas are mistaken) because that isn't deductive.

You need to cite where David Deutsch and Karl Popper say that explanations aren't deductive.


Anonymous at 1:39 PM on July 12, 2018 | #10077 | reply | quote

> When I say I have been pretty clear. This is a conjecture which is not criticised by saying "you are not explaining yourself well".

not even the right part of speech, and not a correct English sentence, is not "pretty clear" when someone is saying you are not communicating.

> Instead of this I should probably say that you have yet to criticise the follow-up attempt I have made.

my guess is you haven't read all my comments (i wrote 3 in a row). i did that and await your reply.

> And we can say what logically follows from what in the investigation.

the word for that is "deducible" ("deduce" for a verb, you switched parts of speech again). the concept you want is called "deduction" not "determination".

> If you conjecture P1 and I conjecture P2 as candidates for the truth.

This isn't correct English, it's a sentence fragment. Is writing correct sentences a thing you could start to try doing? Or do you think that's not something you'll be able to achieve, and not even worth trying? Or you already are trying and this is your best attempt to write a correct sentence?

You wrote a large number of sentence fragments while trying to present a complex argument which includes references to fallacies, by name, without links/cites. That isn't going to work. Make sense?

> All of this is is compatible with the conjecture that P2 is true and also accepting P2, tentatively, of course. You don't add anything to the discussion by saying it should be accepted, other than dogmatism.

You are putting words in my mouth without quoting. Can you stop doing that and try again more seriously?


Dagny at 1:43 PM on July 12, 2018 | #10079 | reply | quote

> You need to cite where David Deutsch and Karl Popper say that explanations aren't deductive.

You read their books (which ones) and didn't understand this? Serious question.

I will look up the cites if you claim to have actually read several books and name them.

It's a different situation if you:

1) wildly misunderstood the books

or

2) never read them

so i want to find out which it is before i do the unpaid work of looking up basic things you ought to already know as background knowledge.


Dagny at 1:44 PM on July 12, 2018 | #10081 | reply | quote

Are you aware that in standard English "explanations" aren't deductive? Or are you going to deny that?


Dagny at 1:45 PM on July 12, 2018 | #10082 | reply | quote

>> All of this is is compatible with the conjecture that P2 is true and also accepting P2, tentatively, of course. You don't add anything to the discussion by saying it should be accepted, other than dogmatism.

>You are putting words in my mouth without quoting. Can you stop doing that and try again more seriously?

The context of the discussion is about the theory of decision making on this site. ET has asserted it. I was not trying to accuse you of it personally.

Suppose that two theories have been put forward: P1 and P2. Suppose also that we investigate these two theories and that we halt the investigation at some time, say t. Now, if at time t there is a criticism of P1 but not of P2, then It cannot be deduced from this investigation that P2 is true or should be accepted. This would only follow if the claims 'P1, 'P2' could not be false together.


Anonymous at 2:08 PM on July 12, 2018 | #10084 | reply | quote

> The context of the discussion is about the theory of decision making on this site. ET has asserted it. I was not trying to accuse you of it personally.

OK but i'm not going to be sympathetic to criticism of ET's ideas without quotes, either. so try again.

waiting on replies to the other stuff.


Dagny at 2:10 PM on July 12, 2018 | #10085 | reply | quote

I have requested citations and your conjectures for my mystification are not exhaustive, but loaded.

If you don't want to give the cites that's up to you.


Anonymous at 2:11 PM on July 12, 2018 | #10086 | reply | quote

> if at time t there is a criticism of P1 but not of P2, then It cannot be deduced from this investigation that P2 is true or should be accepted.

this depends on what acceptance means. you haven't tried to address what you or ET means.

> If you don't want to give the cites that's up to you.

I responded to this and you haven't replied.

Still waiting on replies.


Dagny at 2:14 PM on July 12, 2018 | #10087 | reply | quote

>Are you aware that in standard English "explanations" aren't deductive? Or are you going to deny that?

You mean that the definition of "explanation" does not include the criterion that explanations should be deductive?

I was not using the dictionary definition.

>"Thus an explanation is always the deduction of the explicandum from certain premises, to be called the explicans."

This is from Poppers Objective Knowledge, p.350, 1979.

Notice the word "always" and the word "deduction".


Anonymous at 2:26 PM on July 12, 2018 | #10088 | reply | quote

> You mean that the definition of "explanation" does not include the criterion that explanations should be deductive?

Where does "should" come into it?

DD in FoR terminology section:

> **explanation** (roughly) A statement about the nature of things and the reasons for things.

Popper contradicts himself sometimes. I'll look since you gave a notable quote.


Dagny at 2:29 PM on July 12, 2018 | #10089 | reply | quote

>if the last surviving idea says "you can solve Y problem by Z action", and you have no criticism of that, and you want to solve Y problem, then you should do Z action.

This is an invalid argument. Should does not follow from want.


Anonymous at 2:33 PM on July 12, 2018 | #10090 | reply | quote

> **explanation** (roughly) A statement about the nature of things and the reasons for things.

This does not contradict the criterion

> >You mean that the definition of "explanation" does not include the criterion that explanations should be deductive?

>Where does "should" come into it?

The should does not come into it. That's what the claim means. It might do or it might not do. It does not include the criterion that it should be deductive.


Anonymous at 2:41 PM on July 12, 2018 | #10091 | reply | quote

OK p350:

> In our case the universal law might be put like this: 'If a rat eats at least eight grains of rat poison it will die within five minutes.' The (singular) initial condition (which is a singular statement) might be: 'This rat ate at least eighteen grains of rat poison, more than five minutes ago.' From these two premisses together we may now indeed deduce that this rat recently has died [that is, our explicandum].

This is both false and silly, in different ways.

The way it's false is you can't deduce that conclusion from those premises. The rat might have died a year ago (not recently), then eaten 18 grains of poison 10 minutes ago (recently). No premise that dead rats can't eat was given.

What's going on is that Popper was relying on things other than deduction.

The way it's silly is (abbreviating a little):

X = rat ate 18 poison

Y = rats die if they eat > 8 poison

Z = rat is dead

Popper says X+Y explains Z. But it doesn't explain why poison kills rats or really anything. X and Y are both assertions. They are claims presented without any explanation of how we know them or why they're correct or how they work.

Popper is transferring the need for explanation from Z to Y via guessing X and crafting a suitable Y. This isn't explanation, it's goalpost shifting.

It'd be ok to have this regress if we were getting somewhere but arbitrary assertions don't get you anywhere. And as far as deductive logic is concerned – forget all our background knowledge – Y is nothing but an arbitrary assertion from the set of all logical possibilities. Popper doesn't explain why he chose Y over alternatives which vary the number or the thing eaten or the activity (like you change eating something to sleeping for N minutes within M miles of something). It's not that Popper fails to *fully* explain why he chose Y over its logically possible rivals, it's that he gives *zero* explanation of it. He's relying entirely on unstated appeals to background knowledge, common sense, ... explanations that readers already know, to which he adds nothing.

> Thus an explanation is always the deduction of the *explicandum* from certain premises, to be called the *explicans*.

This doesn't say what an explanation *is*. It says one (alleged) attribute of all explanations.

So my question before addressing this is: What other criterion do you think there are for something to qualify as an explanation?

Popper seems to work with explanations in a common sense (and standard English) way (as discussed in my first section), contrary to this quote.


Dagny at 3:03 PM on July 12, 2018 | #10093 | reply | quote

>> if the last surviving idea says "you can solve Y problem by Z action", and you have no criticism of that, and you want to solve Y problem, then you should do Z action.

> This is an invalid argument. Should does not follow from want.

The problem with this comment is your English fluency or your logical skill.

If I can persuade you of this claim, will you change your attitude to discussion? Will you appreciate it and thank me? Will you recognize you've been making tons of errors and need to stop overreaching?


Dagny at 3:05 PM on July 12, 2018 | #10094 | reply | quote

>>> You mean that the definition of "explanation" does not include the criterion that explanations should be deductive?

>> Where does "should" come into it?

> The should does not come into it.

Then why did you write "should"? (This seems like another major confusion due to your lack of English fluency.)

And could you start trying to use correct English instead of making mistakes constantly? ("The should"?) Or if you can't do that, can you admit it and discuss what to do about it?


Dagny at 3:07 PM on July 12, 2018 | #10095 | reply | quote

>In standard English, this is correct. But with the definition you now claim to be using, it's false. You can't deduce the falsity of theories given proposals (candidate theories to solve a problem), criticism and evidence. Also I don't know what you think "criticism" is, but you can't mean what KP/ET/DD mean (explanations of how/why ideas are mistaken) because that isn't deductive.

This is really confused. You can deduce from a true criticism that something is false. It's called Modus Tolens (MT) .

Suppose you have a theory S, and you come across T which implies ~S. Then by MT & T you can deduce ~S.

I am not saying that the criticism of a proposal is the criticism of the theory.

You seem to be claiming the no explanations are deductive.


Anonymous at 3:10 PM on July 12, 2018 | #10096 | reply | quote

> You can deduce from a true criticism

You can't just assume a criticism is true.


Dagny at 3:20 PM on July 12, 2018 | #10098 | reply | quote

Your post at #10093 does not criticise the claim that all explanations should be deductive. It simply analysises a specific example of an explanation and says what might be in error about it. Just because there are alternative explanations that he did not rule out, does not stop the example from being an explanation.

You are not sufficiently brappling with the issue at hand but bringing up superfluous problems about adequacy of explanation and the problem of other other criteria. Which is not the issue.


Anonymous at 3:22 PM on July 12, 2018 | #10099 | reply | quote

>You can't just assume a criticism is true.

Why not?


Anonymous at 3:23 PM on July 12, 2018 | #10100 | reply | quote

> Your post at #10093 does not criticise the claim that all explanations should be deductive.

It presented relevant discussion and asked you a question before continuing. You ignored the question.


Dagny at 3:23 PM on July 12, 2018 | #10101 | reply | quote

> >You can't just assume a criticism is true.

> Why not?

The context is a deductive proof. Assuming something is true is not a valid step in deductive proof. (If it were, anything could be deduced simply by assuming it's true as a one step deductive proof.)

You have to either start with a premise that it's true, or say "*if* it's true, then..."


Dagny at 3:25 PM on July 12, 2018 | #10102 | reply | quote

You don't understand logic, you can assume things so long as you can discharge it later, Modus Tolens and Modus Ponens are discharge rules.


Anonymous at 3:26 PM on July 12, 2018 | #10103 | reply | quote

> You don't understand logic, you can assume things so long as you can discharge it later, Modus Tolens and Modus Ponens are discharge rules.

1) you're using the word "discharge" wrong and not explaining what you mean

2) why don't you reply to #10094 and we'll find out who doesn't understand logic?


Anonymous at 3:27 PM on July 12, 2018 | #10104 | reply | quote

What warrants a claim to be true? If you can't assume or conjecture it's true. How do you label it true?


Anonymous at 3:35 PM on July 12, 2018 | #10105 | reply | quote

> What warrants a claim to be true? If you can't assume or conjecture it's true. How do you label it true?

you don't label stuff true.

what CR says to do is deduce that X and something else (the criticism, which commonly involves some evidence) contradict (given some background knowledge). that means at least one of them is false. but you can't deduce the falsity (let alone truth) of either one.

Again, please respond to #10094

I haven't made a logical error and you've made lots. You disagree. Let's pursue the clearest example.


Dagny at 3:37 PM on July 12, 2018 | #10106 | reply | quote

>you don't label stuff true.

Why not?


Anonymous at 3:40 PM on July 12, 2018 | #10107 | reply | quote

>> you don't label stuff true.

> Why not?

Well, CR doesn't, and CR explains that we're fallible. Fallible means we can never establish anything is true, so there's no way to label stuff as true.

Again you ignore what I'm say.

Will you stand by #10090 on further examination, or can you find any error in it? That is, did you speak carelessly and you can see the problem now, or are you just unable to find any problem after taking your time?

You challenged my logical skill and I answered you. But you refuse to discuss the massive logical error I think you made. You just won't acknowledge having ever said it.


Dagny at 3:44 PM on July 12, 2018 | #10108 | reply | quote

"Well, CR doesn't, and CR explains that we're fallible. Fallible means we can never establish anything is true, so there's no way to label stuff as true"

So your claim is that only things that can be established to be true can be labelled true.

That is justificationism. Which claims that you can only label stuff true if it's established. CR ruthlessly discards this justificationist assumption and claims that we can conjecture, ie label things true tentatively. So there is a way to label things as true, by conjecture. And then we look for errors. truth is not beyond conjecture. Because I could conjecture something to be true and it be true.


Anonymous at 3:48 PM on July 12, 2018 | #10109 | reply | quote

>You challenged my logical skill and I answered you. But you refuse to discuss the massive logical error I think you made. You just won't acknowledge having ever said it.

Why are you looking for permission to criticise me?


Anonymous at 3:49 PM on July 12, 2018 | #10110 | reply | quote

conjecturing something is true is *not* labelling it true. again your lack of English knowledge is causing confusion. and you jump straight from your lack of English fluency to accusing me of errors (justificationism this time).

I'm done replying until you address #10094


Dagny at 3:50 PM on July 12, 2018 | #10111 | reply | quote

>conjecturing something is true is *not* labelling it true.

And you did not follow it up with an explanation. Which implies that you're quibbling over terminology. We can label something true by conjecturing that it's true. We can't label something true if we require it to be established.

"Label" is neutral here.


Anonymous at 4:05 PM on July 12, 2018 | #10112 | reply | quote

Want to discuss this? Join my forum.

(Due to multi-year, sustained harassment from David Deutsch and his fans, commenting here requires an account. Accounts are not publicly available. Discussion info.)